基于人眼视觉特性的无人机红外图像增强研究  

Infrared image enhancement of UAV based on human visual characteristics

作　　者：陈花卫 谭文斌 CHEN Huawei;TAN Wenbin(Department of Mathematics and Statistics of Big Data Institute,Tongren University,Tongren Guizhou 554300,China)

机构地区：[1]铜仁学院大数据学院,贵州铜仁554300

出　　处：《激光杂志》2020年第9期137-141,共5页Laser Journal

基　　金：贵州省教育厅自然科学研究项目(黔教合KY字[2016]294)。

摘　　要：线性灰度变换方法增强无人机红外图像时,忽略了人眼辨别图像灰度的能力,增强后的红外图像辨识效果差,为此,提出基于人眼视觉特性的无人机红外图像增强方法。基于人眼辨别灰度的特性,在高低灰度区域拉伸灰度间隔、中间灰度区域缩短较大灰度级间隔,由此构建非线性灰度变换模型调整图像灰度;基于对数隶属度函数变换红外图像转至模糊域,对红外图像反复做非线性变换,增强图像边缘信息;基于加权平均融合法融合2种方法增强后的红外图像,得到最终的无人机红外图像增强结果。经测试,该方法增强后的红外图像灰度、亮度适中、边缘效果增强,对比度增益增幅较大。When the linear gray transformation method enhances the infrared image of UAV,it ignores the ability of human eyes to distinguish the image gray scale,and the enhanced infrared image identification effect is poor.Therefore,the infrared image enhancement method of UAV based on human eye vision characteristics is proposed.Based on the characteristics of gray scale discrimination by human eyes,the gray interval in high and low gray areas is stretched,and the large gray interval in the middle gray area is shortened,so as to construct a nonlinear gray transformation model to adjust the image gray scale.The infrared image is transformed to fuzzy domain based on logarithmic membership function,and the nonlinear transformation is repeated to enhance image edge information.Based on weighted average fusion method,the enhanced infrared images of the two methods are fused to obtain the final UAV infrared image enhancement results.The results show that the enhanced infrared image has the advantages of moderate gray scale,moderate brightness,enhanced edge effect and increased contrast gain.

关 键 词：人眼视觉特性 无人机 非线性 灰度变换 红外图像 融合 增强 

分 类 号：TN216[电子电信—物理电子学]